Oscillatory shear flow behaviour of a thermotropic liquid-crystalline polymer

Oscillatory shear flow behaviour of a thermotropic liquid-crystalline polymer Seung Su Kim and Chang Dae Han* Department of Chemical Engineering, Polytechnic University. Brooklyn, NY 11201, USA (Received 23 December 1992; revised 13 April 1993) The oscillatory shear flow properties, dynamic storage and loss moduli (G' and G"), of a thermotropic liquid-crystalline polymer (TLCP) in both the isotropic and nematic regions were measured as functions of angular frequency (~o), using a cone-and-plate rheometer. For the study, an aromatic polyester, poly[(phenyl sulfonyl)-p-phenylene 1,10-decamethylene-bis(4-oxybenzoate)] (PSHQ10), was synthesized in our laboratory. The PSHQ10 was found to have (1) a glass transition temperature of 88°C, (2) a melting point of 115°C, and (3) a nematic-to-isotropic transition temperature of 175°C. We found that the oscillatory shear flow properties of PSHQ10 in the nematic region were strongly dependent on its thermal history (i.e. annealing temperature and the duration of annealing) and shear history. Thus, in order to completely erase the thermal history associated with polymerization and sample preparation, a solvent-cast PSHQ10 specimen was first heated to the isotropic region (e.g. 190°C), sheared there at a very low shear rate (0.0085 s- 1) for ~ 5 min, and then cooled very slowly down to a predetermined temperature (130, 140, 150, 160 or 170°C) in the nematic region. We found that in the isotropic region time-temperature superposition holds and plots of log G' versus log G" give rise to a temperature-independent correlation. On the other hand, we found that in the nematic region preshearing has a profound influence on the oscillatory shear flow properties and log G' versus log G" plots show temperature dependency. This suggests that time-temperature superposition would not hold in the nematic region of PSHQ10 (and also of other types of TLCPs). We thus conclude that log G' versus log G" plots are very sensitive to variations in the morphological state of PSHQ10 as the temperature varies in the nematic region. We found that the Cox-Merz rule holds for PSHQ10 in the isotropic region and the morphological state of a specimen plays an important role in determining whether or not the Cox-Merz rule holds in the nematic region; specifically an unsheared specimen does not follow the Cox-Merz rule, whereas a sheared specimen does. (Keywords: oscillatory shear flow; storage modulus; loss modulus)

INTRODUCTION A better understanding of the rheological behaviour of thermotropic liquid-crystalline polymers (TLCPs) is essential for the successful processing of such polymers, In recent years, severalresearch groups 1-15 have reported on the transient and/or steady shear flow behaviour of certain TLCPs. However, relatively few studies 2'~5-1a have been reported of the oscillatory shear flow behaviour of TLCPs. In the investigation of the rheological behaviour of TLCPs, measurements of oscillatory shear flow properties have certain advantages over measurements of steady shear flow properties; specifically, (1) the time evolution of the morphological state in the nematic region of a specimen, after being subjected to transient or steady shear flow, can be monitored in terms of the dynamic storage modulus (G') or complex viscosity (1~*1), by applying small amplitude oscillatory deformations, and (2) oscillatory shear flow measurements allow us to apply much higher angular frequencies, as compared to steady shear flow measurements which are limited to low shear rates when dealing with very viscous molten polymers, *To whom correspondence should be addressed at Department of Polymer Engineering, The University of Akron, Sidney L. Olson

Research Center, Akron, OH 44325-0301,USA 0032-3861/94/010093-11 ,~ 1994 Butterworth-Heinemann Ltd.

Very recently we have embarked on an investigation of the rheological behaviour of TLCPs. In this endeavour, we have synthesized a thermotropic polyester, poly[(phenyl sulfonyl)-p-phenylene 1,10-decamethylenebis(4-oxybenzoate)] (PSHQ10), with the chemical structure19: [

-

~

-]

T~'--~ J / ~ C ~ o ~ O - - c ~ o - - ( C H 2 ) I 0 L ~ ' / OII ~ II ~x~/ O

/

In

so2

©

PSHQIO

Previously, Driscoll and co-workers 15'2°'2~ reported the rheological behaviour o f a thermotropic polyester having a chemical structure similar to that of PSHQ10, which had been synthesized by Furukawa and Lenz ~9. It should be mentioned that the number 10 in PSHQ10 refers to the numberofmethyleneunitspresent, i.e. decamethylene or (CH2)~o. This convention is chosen in order to distinguish PSHQs with varying numbers of methylene units as flexible spacers. In a future study we intend to synthesize PSHQs with varying lengths of flexible spacers, e.g. PSHQ8 with (CH2)a and PSHQ12 with (CH2)12.

POLYMER, 1994, Volume 35, Number 1

93

Oscillatory shear flow behaviour." S. S. Kim and C. D. Han In the investigation of the rheological behaviour of TLCPs, PSHQ10 has some unique features, namely (1) PSHQ10 is a homopolyester, thus allowing us to offer a much simpler interpretation of rheology-morphology relationships, as compared to the situations which deal with copolyesters, such as copolyesters ofhydroxybenzoic acid (HBA) and poly(ethylene terephthalate) (PET), and copolyesters of HBA and 2-hydroxy-6-naphthoic acid (HNA); (2) PSHQ10 has a relatively low nematicto-isotropic transition temperature (TM= 175°C, also referred to as the clearing or isotropization temperature), which is significantly lower than the thermal degradation temperature (~ 350°C), thus allowing us to take rheological measurements in both the isotropic and nematic regions; (3) PSHQ10 has a 60°C range between the melting point (115°C) and TN=(175°C), thus enabling us to take rheological measurements over a reasonably wide range of temperatures in the nematic region. Owing to the fact that the TN~of PSHQ10 is much lower than the thermal degradation temperature, in our previous study 22 we have shown how the initial morphology of PSHQI0 in the nematic region can be controlled, which is very important for the investigation of transient shear flow behaviour, In this paper we shall present the results of our recent study on the oscillatory shear flow behaviour of PSHQ10 in both the isotropic and nematic regions. Emphasis will be placed on pointing out that both thermal and shear histories have a profound influence on the oscillatory shear flowbehaviourofPSHQ10inthenematicregion. EXPERIMENTAL

Polymer synthesis In the present study PSHQ10 was synthesized via solution polymerization, in our laboratory, by preparing two monomers, 2-phenylsulfonyl-l,4-hydroquinone (A) and 4,4'-dichloroformyl-l,10-diphenoxydecane (B), with the chemical structures19: __~ / ~ - k s°2

H~OH 2-phenylsulfonyl- i, 4-hydroquinone

(A)

Cl--C---~-------~O--(CH2)lo--O~~-~ C--Cl II ~ / ~ / I I o

o

The preparation of monomer B required three steps. (1) 4,4'-Dicarbomethoxy-l,10-diphenoxydecane was prepared by reacting methyl-4-hydroxybenzoate dissolved in N,N-dimethylformamide (DMF) with 1,10dibromodecane in the presence of anhydrous sodium carbonate. The reaction took place with stirring for 4 h in a 2 litre three-necked flask, which was immersed in an oil bath maintained at 135°C. The reaction mixture was cooled and poured into cold water, and then placed in a refrigerator overnight. The crude material was later filtered and washed thoroughly with water and dried. (2) 4,4'-Dicarboxy-l,10-diphenoxydecane was prepared by pouring the reaction product obtained in step 1 into a 10wt% solution of potassium hydroxide in ethanol. The resulting suspension was stirred with reflux for 4 h and then diluted with water and, finally, heated to 90°C, thus dissolving the potassium salt in the product. Hydrochloric acid solution was added to the solution containing the reaction product until the mixture became distinctly acidic. The solution was allowed to cool overnight before being filtered. The crude product was washed with water and recrystallized several times from a DMF/ethanol mixture. (3) Due to the air-sensitivity of the chlorinated compound, monomer B, chlorination of the product obtained in step 2 (hereafter referred to as Pre-B) was performed immediately before the polymerization to avoid chemical changes of functional groups in both ends of Pre-B during storage. A precise amount of Pre-B was refluxed with thionyl chloride for 4h, yielding monomer B. Excess thionyl chloride was then removed under reduced pressure for several days. Monomers A and B in equimolar quantities were stirred in a mixture of dichloromethane and pyridine. Moisture-free and oxygen-free argon gas was kept flowing into the reactor for 2 days. The reactor was then sealed and stirred for 2 days to obtain the maximum degree of polymerization. The polymer was fractionated by precipitating several times in a mixture of methanol and dichloromethane. High molecular weight fractions precipitated in the form of white fibrils and low molecular weight fractions were obtained in the form of a fine powder. A fibrous product was recovered by filtration using a coarse screen followed by washing several times with a methanol/water mixture and drying in a vacuum oven for several days.

Characterization of PSHQIO The chemical structure of PSHQ10 was confirmed via two-dimensional high-resolution n.m.r, spectroscopy. The samples were prepared in CDCI 3 and spectra were acquired at room temperature on a Varian Instruments Gemini 200 n.m.r, spectrometer operating at 200 MHz for 1H and 50 MHz for 13C. The n.m.r, peak assignments

4,4'-dichloroformyl-l, 10-diphenoxydecane (B)

for P S H Q I 0 are summarized in Table 1. The 1H C O S Y was acquired as a 256 x 512 data set and zero filled to

Monomer A was prepared by reacting at room temperature a solution of an excess amount of sodium benzenesulfinate in a 5:1 mixture of water and acetic acid with a solution of p-benzoquinone (purified by sublimation) in a 5:1 mixture of water and acetic acid. The crude product, precipitated in white powdery form, was filtered and washed with water several times, and then dissolved in methanol and recrystallized from a methanol/water mixture several times. After being dried in vacuo, the product was then recrystallized from an ethanol/water mixture several times,

512 x 512 real points. It was apodized with a sine bell in both dimensions. The COSY was symmetrized after inspection of the data to assure no complications from Tx noise. Since the major limitation in the ~H COSY is that there are so many unprotonated carbons in the molecule, we used ~3C n.m.r, which would observe all 20 of the different types of carbon. The ~3C chemical shifts are characteristic of the structure, but do not provide sufficient information to confirm the structure since a number of the peaks are quite close to each other, e.g. f and s (Table 1). Therefore we used a two-dimensional

94 POLYMER, 1994, Volume 35, Number 1

Oscillatory shear flow behaviour. S. S. Kim and C. D. Hun Table 1

N.m.r. peak assignments for P S Q H I O

0 II

-4CHzs~ff=~Ci-f g

0 o n II ~ r J ~ - ~ 9 - O--Cp ~ ' SO2 A c c

s

O--(CHz~)-ff-

b~k-~b

Assignment

a ~H

13C

a b

7.75 7.30

133.90 129.40

c e f g h i j k 1 d

m n o p

q r s t

7.80 7.05 8.16

-

-

7.52 7.24 8.15 -

8.01

6.98 -

128.50 164.40 114.98 132.97 121.12 164.87 148.70 128.90 134.55 140.81

146.30 126.45 123.75 163.90

120.85 133.21 114.80 164.40

heteronuclear cross correlation (HETCOR) experiment to determine which protons are on which carbons. The H E T C O R was acquired as a 128 x 2K data set, used Gaussian apodization in both dimensions and it was zero filled to 256 in T1. This experiment is analogous to the 1H COSY, but since the axes are not the same there are no 'diagonal' correlations. Rather, contours appear where the 1H and ~3C resonances are J-coupled to one another, i.e. a specific C-H pair. To summarize, by combining the 1H and ~3C chemical shift information, the COSY homonuclear ~H information on neighbouring groups and the H E T C O R information which tells us which protonated carbons are neighbours we were able to assign the entire spectrum. The results are consistent with the proposed PSHQ10 structure. We have determined the weight-average molecular weight (Mw) of PSHQ10 using g.p.c. In the present study we employed two different PSHQ10s, one having M~, of 45000 relative to polystyrene standards and a polydispersity index of 2, and the other having a Mw of 35000 relative to polystyrene standards and a polydispersity index of 2. Unless stated explicitly, the results presented below are based on the PSHQ10 specimens with Mw = 45 000.

Rheological measurement A model R16 Weissenberg rheogoniometer (Sangamo Control, I n c . ) i n the cone-and-plate (25 mm diameter plate and 4 ° cone angle)configuration was used to measure, (1) in the steady-state shear mode, the shear stress (a) as a function of shear rate (7), and (2) in the oscillatory shear mode, the dynamic storage modulus (G') and dynamic loss modulus (G") as a function of angular frequency (co), under isothermal conditions. From the measurements of G' and G", the absolute value of complex viscosity was calculated using the relationship: [/7*(O))l = { r a ' ( o ) ) / ( o l 2 .~ [ G " ( o ) ) / o ) ] 2 } 1/2

Strain amplitude was varied from 0.01 to 0.06, which was well within the linear viscoelastic range of the materials investigated. This, together with data acquisition during measurement, was accomplished with the aid of a microcomputer interfaced with the rheometer. All experiments were conducted in the presence of nitrogen in order to preclude oxidative degradation of the specimen. Temperature control was satisfactory to within + loc.

Thermal transitions of PSHQIO Thermal transition temperatures of the synthesized PSHQ10 were determined by d.s.c. (Perkin-Elmer DSC7). Figure 1 shows a d.s.c, trace during the first heating cycle for a solvent-cast PSHQ10 specimen with Mw = 45 000, which was annealed at 160°C for 20 h. D.s.c. runs were made under a nitrogen atmosphere with heating and cooling rates of 20°C m i n - 1. It can be seen from Figure 1 that PSHQ10 has (1) a glass transition transition of ~ 88°C, (2) a melting point of ~ 115°C, and (3) a TN~ of ~ 175°C. Before embarking on the present study, we spent a considerable amount of effort 22 in trying to understand the effect of thermal history on the rheological behaviour of P S H Q 10. RESULTS AND D I S C U S S I O N

Oscillatoryshearflow properties of PSHQIO in the isotropicregion The oscillatory shear flow properties of PSHQ10 in the isotropic region were measured at 180, 185 and 192°C, and are summarized in terms of reduced

~z5

/

~

___....-

8

o

Sample preparation Specimens for rheological measurements were prepared by first dissolving PSHQ10 in dichloromethane as solvent in the presence of an antioxidant (Irganox 1010, Ciba-Geigy Group) and then slowly evaporating the solvent at room temperature for 1 week. The cast films (1 mm thick) were further dried in a vacuum oven at room temperature for at least 3 weeks and at 90°C for 48 h. Immediately before the rheological measurement, the specimen with Mw = 4 5 000 was further dried at 120°C for 2 h to remove any residual solvent and moisture,

~

~

8

~

~/~

50

I 7~

L ioo

I iz5

I Lso

I 175

I aoo

aa5

Temperature (°C) Figure 1 D.s.c. trace for a solvent-cast P S H Q 1 0 specimen, which was annealed at 160°C for 20h, showing that (1) the glass transition temperature is 88°C, (2) the melting temperature is 115°C, and (3) the nematic-to-isotropic transition temperature is 175°C

POLYMER, 1994, Volume 35, Number 1

95

Oscillatory shear flow behaviour. S. S. Kim and C. D. Han ~o5 _

_i,~,~

, ~ -

//

104

,¢,~¢

"~ ,~ o 1°3---

~ ~'-~.

~ 104 no ~ %

~---7 O~

! io~

temperature-independent correlation. On the basis of the previous studies of Han and co-workers 24-29, we can conclude from Figure 3 that the data were indeed taken in the isotropic region. This then assures us that the time-temperature superposition used in Figure 2 is justified. When discussing the oscillatory shear flow properties of PSHQ10 in the nematic region below, we shall make use of the rheological significance of Figure 3.

~o5

__--

/I~1

--_ l03

IZl I11

1 Ilflll! l0z Jd

ill[

fr'

Io° O-)CIT(rod/s)

Figure2 Reduced plots for logG', versus logtoaT at: ((D) 180°C, (A) 185°C, (11) 192°C;log G~' versus log coat at: (O) 180°C, (&) 185°C, (ll) 192°C; and loglq*l versus logcoar at: (G)) 180°C, (A) 185°C, ([]) 192°C for PSHQ10 in the isotropic region. The reference temperature used is 180°C I°5 _ --

-

-

Oscillatory shearflow properties of PSHQIO in the nematic region Effect of thermal history. For this experiment, PSHQ10 with M,~=35000 was employed. In order to investigate how annealing conditions might influence the oscillatory shear flow properties of PSHQ10 in the nematic region, we varied the time of annealing after a temperature PSHQ10. Plots of log specimen wasbelow placedthein TNj the ofcone-and-plate fixture at G'a versus log co are given in Figure 4, and plots of log G" versus log co are given in Figure 5, at 140°C for PSHQ10 specimens having the following thermal histories: (1) a solvent-cast specimen which was annealed at 140°C for 2h; (2) a solvent-cast specimen which was annealed at 140°C for 4.5 h; (3) a solvent-cast specimen which was annealed at140oC for 24 h; and (4) a solvent_cast specimen which was first annealed at 140°C for 24h and then thermally treated at 200°C for 30 min, and finally cooled very slowly down to 140°C for further annealing for a period of 24 h. Note that a fresh specimen was used for each frequency sweep. The seemingly complicated frequency dependencies of G' and G" for PSHQ10 in the nematic region at 140°C, displayed in Figures 4 and 5, can easily be explained by using the well-established frequency dependencies of G' and G" for viscoelastic fluids over a wide range of co

,o4 ~-

io 5 -

_

_-

[]

o

~

E]-

o

L~ ~o3 --

z

/~2

"-

Z

F_

~

104 ----

,o2-

-~

_--

4x101

io3

IIIll

4xlO 2

I

I

I IIIill

I03

I

i

iiiiii

I04

I05

G" (Pa) Figure 3 Plots of log G' versus log G" for PSHQ10 in the isotropic region at: (Q) 192°C; (/3,) 185°C; (Fq) 180°C 102 variables in Figure 2: G'r = G'poTo/pT, G~= G'poTo/pT,

andlrl,l=lrl,lPoTo/pT. Note in Figure 2 that aT is a shift factor and the subscript 0 denotes a reference state. A close look at Figure 2 reveals that although PSHQ10 is a semiflexible polymer, the essential features of the reduced plots are little different from those often observed 23

for flexible macromolecules . Plots of log G' versus log G" for PSHQ10 in the isotropic region are given in Figure 3, showing a

96

P O L Y M E R , 1 9 9 4 , V o l u m e 35, N u m b e r 1

4,1o' t°-t

J J i IJllJl I°°

I I flllrJ

Io'

I r I JlIJI ao2

~ (rad/s) Figure4 P l o t s o f l o g G ' v e r s u s l o g ~ o a t l 4 0 ° C f o r s o l v e n t - c a s t P S H Q l O specimenshaving the followingthermal histories:(Q) annealed at 140°C for 2 h; (~) annealed at 140°C for 4.5 h; (Fq) annealed at 140°C for 24 h; (0) first annealed at 140°C for 24 h and then heated to 200°C, held there for 30min, subsequently cooled slowly down to 140°C and annealedthere for24h

Oscillatory shear flow behaviour: S. S. Kim and C. D. Han

3 x Mo 4 -

Io4 _--

_ % io3 -

102

I

iv-~

I I I I IIll l0°

I ~ I I I r =l] iv~ to Crad/s)

I I t I I~ =oz

Figure 5 Plots of logG" versus logto at 140°C for solvent-cast PSHQ10 specimens having the same thermal histories as the specimens in Figure 4

~og i Terminal

Plateau

i zone

zone

~

..W----~,/ G'

% z, ",~

Transition

d . ~o,e / . ~¢'

~ G" / G'

~og

to Figure 6 Schematic diagram of plots of log G' versus log to and log G" versus log to for viscoelastic polymeric liquids over a wide range of angular frequencies

Specifically, the following observations are worth noting; (1) log G' versus logo9 and log G" versus log o9 plots for the specimens which were annealed at 140°C for 2 and 4.5 h, respectively, lie in the terminal region; (2) log G' versus log o9 and log G" versus log 09 plots for the specimen which received thermal treatment at 200°C and subsequently annealed at 140°C, also lie in the terminal region; (3) log G' versus log 09 and log G" versus log o9 plots for the specimen which was annealed at 140°C for 24 h, lie in the plateau region. The above observations indicate that as annealing continues, the structure of PSHQ10 in the nematic region changes with time, giving rise to different viscoelastic responses. Using the data given in Figures 4 and 5, we prepared the plots of log G' versus log G" and they are given in Figure 7. It can be seen from Figure 7 that the thermal history of a specimen greatly influences not only the shape of log G' versus log G" plots, but also the magnitude of G' for a fixed value of G", namely the value of G' increases with increasing annealing time. Recall that log G' versus log G" plots of PSHQ10 in the isotropic region are independent of temperature (see Figure 3). This then

In other words, log G' versus log G" plots are sensitive to variations in the morphological state of PSHQ10 (and also of TLCPs in general). We can also conclude from Figure 7 that the melt elasticity (i.e. values of G' for a fixed value of G") of a specimen which received thermal treatment in the isotropic region at 200°C is less than that of the specimens which did not receive thermal treatment in the isotropic region. It is appropriate to mention at this juncture that earlier Han and Kim 3° analysed the rheological data of Aoki 31 who reported the oscillatory shear flow properties of poly(acrylonitrile-stat-butadiene-stat-styrene) (ABS) resins containing various amounts of rubber particles, and showed that log G' versus log G" plots were very sensitive to the amounts of rubber particles present in the ABS resins, i.e. the greater the amount of rubber particles in an ABS resin, the larger the values of G' (i.e. the more elastic). Also, when investigating the rheological behaviour of microphase-separated block copolymers, and mixtures of a microphase-separated block copolymer and a homopolymer, Han and coworkers 3°'32-35 observed the temperature dependence of log G' versus log G" plots, very similar to that given in Figure 7. On the basis of these observations, we can conclude from Figure 7 that the solvent-cast PSHQIO specimens, which were subjected to different thermal histories, must have had different morphological states. Plots of log Irt*l versus log o9 are given in Figure 8. The following observations are worth noting: (1) the specimen, which was annealed at 140°C for 2 h, exhibits shearthinning behaviour over the entire range of o9 tested; (2) shear-thinning behaviour becomes more intense as the annealing period increased from 2 to 4.5 h and to 24 h; (3) when a solvent-cast specimen received thermal treatment at 200°C for 30min and subsequently was annealed at 140°C for 24 h, plots of log I~*l versus log o9 exhibit (i) a very weak shear-thinning region at low co, (ii) a Newtonian plateau region at intermediate values of O9, and (iii) a strong shear-thinning region at high o9. Of

Io5

( F i g u r e 6) 23 .

s u g g e s t s t h a t l o g G' v e r s u s l o g G" p l o t s are v e r y sensitive

to the thermal history, which in turns affects the morphology of PSHQ10 specimens in the nematic region,

io* -~ "~ =o~ /( ~_J 2 ivz _ [

~oz

I

iv3

=o4

I

I

I Illl

iv5

G° (Po) Figure 7 Plots of log G' versus log G" at 140°C for solvent-cast PSHQ10specimens, having the same thermal histories as the specimens in Figure 4

P O L Y M E R , 1 9 9 4 , V o l u m e 35, N u m b e r 1

97

Oscillatory shear flow behaviour." S. S. Kim and C. D. Han 4xlO~

behaviour of PSHQ10 are referred to in our previous paper22.

-

Hysteresis. For this experiment, P S H Q 1 0 with Mw=35000 was employed. The measurements of Iq*l by increasing or decreasing 09 for a solvent-cast specimen, which was annealed for 443 h at 150°C, are given in Figure 9, where open circles represent the data taken by increasing o9 stepwise (ascending frequency sweep) and filled circles represent the data taken by decreasing 09 stepwise (descending frequency sweep). It can be seen in Figure 9 that the values of I~/*l for the descending frequency sweep are lower than those for the ascending frequency sweep, indicating that the PSHQ10 specimen exhibits a hysteresis effect in the measurement of oscillatory shear flow properties. Earlier, similar observations were made by Wissbrun 2, who used thermotropic copolyesters. Figure 10 describes the measurements of Iq*r as functions ofo9 for specimens having the following thermal

io"

g

i 1041

io3

lOCI

,o-'

I

I IlllllJ

I

/llllllJ

,o°

I

I IIIII

,o'

io4 ,oz

~(' )

(,0 ( t a d / s )

Figure8 Plots of logJ~/*l

logoJ at 140°C for solvent-cast PSHQ 10 specimens having the same thermal histories as the specimens versus

in Figure 4

particular note in Figure 8 is that at low o9 the specimen which received thermal treatment at a temperature above TN] has much lower values of Ir/*l than the one which did not receive such a thermal treatment in the isotropic region. In other words, the thermally treated specimen appears to exhibit a three-region viscosity curve, observed earlier by Onogi and Asada 36. We believe that the complicated frequency dependence of I~/*1, observed in Figure 8, is attributable to the

differences in the morphology that existed in specimens, which had experienced different thermal histories. Similar observations were reported earlier by Lin and W i n t e r 37, w h o u s e d a thermotropic copolyester o f H B A a n d H N A . It should be mentioned, however, that owing to the fact that the TNI of the HBA/HNA copolyester was above or close to its thermal degradation temperature ( ~ 400°C), in the study of Lin and Winter 37 it was not possible to heat samples to an isotropic state before being subjected to oscillatory shear flow measurements in an anisotropic state. In a previous study 22, using d.s.c, we observed the formation of a 'crystal-like' phase when PSHQ10 was annealed at temperatures in the nematic region between 130°C and 140°C and the disappearance of the 'crystallike' phase when PSHQ10 was first heated to an isotropic state and then subsequently cooled down to an anisotropic state. This then explains why in Figure 8 the [r/*l of the specimen which was annealed at 140°C without receiving thermal treatment at an isotropic state at 200°C, increased with increasing annealing time. On the other hand, the values of Irt*l for the specimen which received thermal treatment at 200°C are lower than those of the specimens that did not receive thermal treatment at a n isotropic state. This seems to suggest that the morphology of the specimen which received thermal t r e a t m e n t at a n isotropic state is quite different from that of the specimens which did not receive such thermal treatment. The details of the effect of thermal history on the rheological

98

O

o ~(2) Ooooo oo oo

POLYMER, 1994, Volume 35, Number 1

(7)

"~ IO~_

--

--

~

-_ ~_ -I 0

(3)

(6)

°oooooooo ~ $

I zo

I 30

I 40

I 50

I 60

•

I 90

J ,00

~,

~ q ~

I 40

°°°°

•~

f zo

soI

Time train)

Figure 9 Hysteresis effect observed in the complex viscosity of a solvent-cast PSHQ10 specimen, which was annealed at 150°C for 443 h: ((3) data taken for ascending frequency sweep; (@) data taken for descending angular frequency sweep. The arrows indicate the time at which the angular frequency was increased for ascending frequency sweep or decreased for descending frequency sweep. ~o (rads-1): (]) 0.24; (2) 2.37; (3) 9.45; (4) 23.7; (5) 9.45; (6) 2.37; (7) 0.24

104_

~' i03 g -~

I°2 _ ~o-'

I I J ~J~t~l io0

J J J ~,,rrl io~

I I I II I io2

w (rod/s) Figure 10 Plots of loglq*l v e r s u s l o g o for solvent-cast PSHQ10 specimens having the following thermal histories: (Q, O) first annealed at 140°c for 3 h and then heated to 150°C for further annealing for 80 h; (z2x, &) thermally treated at 190°C for 30 rain and then cooled slowly down to 150°C. Open symbols represent the data taken for the ascending frequency sweep, and filled symbols represent the data taken for the descending frequency sweep

Oscillatory shear f l o w behaviour: S. S. Kim and C. D. Han

histories: (1) a solvent-cast specimen was first annealed at 140°C for 3 h and then at 150°C for 80 h, and (2) a solvent-cast specimen was first heated into the isotropic region at 190°C for 30 min and then cooled very slowly down to the nematic region at 150°C. It can be seen from Figure 10 that, as compared to the specimen which was annealed into the nematic region at 140°C, the specimen which first received thermal treatment in the isotropic region at 190°C has: (i) lower values of Ir/*l, (ii) less hysteresis effect, and (iii) less intense shear-thinning behaviour. From the above observations we conclude that the thermal history of a PSHQ10 specimen has a profound influence on the extent of the hysteresis effect in its rheological behaviour, Control of initial morphology. In order to control the initial morphology of PSHQ10 in the nematic region, in the present study the following experimental procedures were adopted. (1) First, a solvent-cast PSHQ10 specimen was placed in the cone-and-plate fixture at 190°C, and after the temperature was equilibrated a steady shear flow at a rate of 0.085 s-1 was applied for 5 min to the specimen. Oscillatory shear flow measurements were then taken over a wide range of e) at 190°C. (2) The specimen was then cooled very slowly down to a predetermined temperature (130, 140, 150, 160 or 170°C) in the nematic region. It took ,~ 1.5 h to cool the specimen from 190 to 140°C (there was no convective cooling device installed in the instrument). (3) After the temperature was equilibrated in the nematic region, the specimen was subjected to oscillatory shear flow over a wide range of ~o. Figure 11 gives plots of log G' versus logo), and Figure 12 gives plots of log G" versus log 09, for PSHQ10 specimens at 130, 140, 150 and 160°C, respectively, after each of the specimens was cooled from the isotropic region. Note that a fresh specimen was used for each temperature. It can be seen in Figure 11 that the shape of the log G' versus log 09 plot very much resembles that for microphase-separated block copolymers 3°'32'33, in that curvature is seen in the terminal region. This appears to suggest that at temperatures ranging from 130 to 160°C PSHQ10 is no longer a homogeneous polymer, which is

io5

1°4 g_ "b I°'~ 4o~ I t i itltll J tlJlliLI J I Ll IJill I I t lIHIl J I t rl, Lo-3 Io-2 Io-~ IO° IO~ io2 w (rod/s) Figure 12 Plots of logG" versus logto for solvent-cast PSHQ10 specimens, which have the same thermal histories as the specimens in Figure 11

1°5=

I°4 -

-~ ~ iJ _-"o --

j,

I

I I I IIIll

I

I

I I llilI

I

104

,03

G"

I 4xlO4

(Po)

Figure 13 Plots of log G' versus log G" for solvent-cast PSHQ10 specimens, which have the same thermal histories as the specimens in Figure 11, except for (Q) which represents the data taken in the isotropic region at 190°C

104

1°3

"~ io2 = -

IO I

l

/

I°z E-

10 2

v

~

-

i01

105

,

I I IIIIlll

Io-3

Io-a

I I lllllll I II IIIIIl I J~lm IO-I IO° IOI ro2 oJ (rod/s)

Figure 11 Plots of log G' versus log~o for solvent-cast PSHQ10 specimens after being thermally treated in the isotropic region at 190°C and then cooled slowly down to the nematic region at: (O) 130°C; (a) 140°C; (I) 150°C;(V) 160°C

expected to have a slope of 2 in the terminal region 23. On the other hand, the shape of the log G" versus log plot given in Figure 12 does not show any unusual features which would be regarded as being distinctly different from those for homogeneous polymers. It should be mentioned that the specimens (Mw = 35 000) used in Figures 11 and 12 had thermal histories different from those used in Figures 4, 5, 7 and 8. Using the data given in Figures 11 and 12, we have constructed the plots of log G' versus log G" displayed in Figure 13, where for comparison plots of log G' versus log G" for PSHQ10 in the isotropic region are also given. It is of interest to observe in Figure 13 that log G' versus log G" plots show temperature dependence, indicating that the morphology of PSHQ10 varies with temperature.

POLYMER, 1994, Volume 35, Number 1

99

Oscillatory shear flow behaviour." S. S. Kim and C. D. Han

lore z

-"

-

1°5--

~ . ~ _ e ' e ' ~ "e , = ~

-

104---

-

-I Io~

v-~"~--~-~-.1,.qL_ ~

"

~

~o2/I ~H~,I J l lllJ~il J a iJlJlJl i i i liHd i i ~J~u~ io-a io-2 io_l I°o I°~ I°2 t.o (rod/s) Figure 14 Plots of log I~*1 v e r s u s log to for solvent-cast specimens,

which have the same thermal histories as the specimensin Figure 11

Note in Figure 13 that as the temperature is increased from 130 to 160°C, log G' versus log G" plots move towards those in the isotropic region. Such temperature dependence of log G' versus log G" plots has been observed for microphase-separated block copolymers by Han and co-workers 3°'32-35. It can then be concluded that log G' versus log G" plots would be useful to determine the TN] of TLCPs. What is significant about Figure 13 is that log G' versus log G" plots are very sensitive to any variation in the structure of the polymer and thus can be used to monitor the initial morphology of the PSHQ10. In a separate study 3s, we indeed used log G' versus log G" plots to monitor the initial morphology of PSHQ10 when conducting transient shear flow experiments, Figure 14 gives plots of log l~/*l versus logo9 at 130, 140, 150 and 160°C for PSHQ10 specimens (M, = 35 000) after being cooled down from the isotropic region. Note that a fresh specimen was used for each temperature. We observe in Figure 14 that at 130°C PSHQ10 exhibits shear-thinning behaviour over the entire range of co tested, but the extent of shear-thinning behaviour becomes less as the temperature increases, and at 160°C a Newtonian plateau region appears at low co and a shear-thinning region at higher o9. Note that Figure 14 was prepared using the data given in Figures 11 and 12. The shape of the curve in Figure 14, especially at low values of co, is affected much more by the magnitude of G" than by the magnitude of G' because, according to Figures 11 and 12, at low values of co the magnitude of G" is much greater than that of G'. In view of the fact that the shape of the log G" versus log to plot, for instance, at 160°C given in Figure 12 is very similar to that for homopolymers with flexible chains, it is not surprising to observe in Figure 14 that the log 1,7"1versus log co plot at 160°C very much resembles that for homopolymers with flexible chains. On the other hand, from the micrographs taken in our previous study 22 we know that PSHQ10 at 160°C has a liquid-crystalline phase in the form of a polydomain structure. We can thus conclude that the use of the log G' versus log G" plot is much more effective than the use of the log I~/*lversus log co plot, in determining whether a LCP is in the isotropic region or in the anisotropic region,

100

POLYMER, 1994, Volume 35, Number 1

Temperature dependence ofl~/*l. Figure 15 gives plots of log I~/*1 versus temperature for a PSHQ10 specimen at o9=0.24 rad s-1. Note that the data points in Figure 15 were obtained for specimens before being subjected to steady shear flow in the nematic region, and a sufficiently long time was allowed to attain temperature equilibrium. It can be seen in Figure 15 that Ir/*l goes through a maximum at ~175°C, which is very close to the TN] determined by d.s.c, under quiescent conditions, and that Ir/*lat 170°C is higher than that at 160°C. In reference to Figure 15, the temperature dependence of I~/*1can be divided into three regions: (1) the nematic region, where T< 165°C and I~/*1decreases with increasing temperature; (2) the biphasic region, where 165°C < T< 175°C and I~*1 increases with increasing temperature; and (3) the isotropic region, where T > 175°C and Ir/*[ decreases with increasing temperature. While the temperature dependence of I~/*1 in the isotropic region can be explained by the Arrhenius rule, the interpretation of the temperature dependence of I~*1 in the nematic region is not as simple, due to the fact that the morphology of PSHQ10 varies with temperature. The increase of I~/*1in the biphasic region is attributable as being due to the polydisperse nature of PSHQ10 (having a polydispersity index of ~ 2), since the low molecular weight portion of the specimen would begin to transform to the isotropic phase before reaching the TNI, 175°C. Note that the viscosity of a TLCP is greater in the isotropic region than in the anisotropic region, because the anisotropic phase undergoes orientation in the shear direction. Therefore as the temperature is increased from ~ 165 to 175°C, the amount of isotropic phase in the bisphasic region will increase accordingly, thus increasing the weight-averaged viscosity of the mixture. It should be mentioned that the biphasic nature, as investigated by the rheological method, of other TLCPs was reported by other investigators TM1. Effect of steady shear flow. Let us now discuss how the application of steady shear flow might affect subsequent oscillatory shear flow properties of PSHQ10 in the nematic region. It should be remembered that in the present study the specimen, which was subjected

Zx=°5["

/

IoSI--

~_ NN~x \ ~ g_ _Z ~

~ ~

Nemotic region Isotropic

.

io4~

Biphosic//-~SL.rag,on " ~

~

t

2 x Io3

130

I

140

I

150

I

160

I

170

Temperature

I

180

I

190

200

(%)

Figure 15 Temperature dependence of I~*1for PSHQ10 at (o=0.24

rad s-1

Oscillatory shear flow behaviour. S. S. Kim and C. D. Hun to steady shear flow followed by oscillatory shear flow in the nematic region, had already received thermal treatment in the isotropic region and had been cooled to the nematic region. In the present study we measured the oscillatory shear flow properties of PSHQ10 specimens after they had been allowed to rest for a fixed period upon cessation of steady shear flow. During the rest period, reorganization or reformation of domain textures (often referred to as 'structural recovery') would take place in the specimens. It should be mentioned that in the past, small amplitude oscillatory deformations were used to monitor the time evolution of structural recovery, in terms of complex viscosity It/*l under small amplitude oscillatory deformations, in lyotropic or thermotropic LCPs after cessation of transient or steady shear flow 14'39. Such experiments were conducted on the premise that the morphology of the specimen would not be altered by the small amplitude oscillatory deformations applied to the specimen. Very recently we also conducted similar experiments using PSHQ10 and the details of the study will be reported in a future publication. Figure 16 describes the effect of steady shear flow on subsequent oscillatory shear flow property, 1,1"1, for PSHQ10 specimens at 150°C, where a fresh specimen was used for each shear rate. Also given in Figure 16, for comparison, are plots of log I *1 versus log o9 for an unsheared PSHQ10 specimen (i.e. before being subjected to steady shear flow). It can be seen in Figure 16 that the application of steady shear flow decreased the I~/*1of a specimen, and that as the applied shear rate increased, the values of In*l after shearing approached those before shearing, Figure 17 shows log G' versus log G" plots at 150°C for PSHQ10 specimens after being sheared at a shear rate of 0.107, 0.526 and 1.07 s -1. Note that a fresh specimen was used for each shear rate. Also plotted in Figure 17, for comparison, are log G' versus log G" plots at 150°C for a PSHQ 10 specimen before being sheared and for a PSHQ10 specimen in the isotropic region at 190°C. It can be seen that the application of steady shear flow

Lo5

to4 "~ g ~o~

4 x l O4

Io4---

/

/

,.J

flP~

_

\ Lsotropic phase

io.~ _

no -~

-

-

/4 J:z

ioz

~o

) /

./

I i i I i i Ill I t ~o4 4 x io'~ G" (Po) Figure 17 Effectof steadyshear flowon subsequent oscillatoryshear flow properties, shown by log G' v e r s u s log G" plots, for PSHQ10 specimensat 150°C.The shear histories are the same as the specimens in Figure 16. Symbolshave the same meaningsas in Figure 16, except for (Q) whichrepresentsthe data takenin the isotropicregionat 190°C lo2

I I I I I I Ill

1o3

shifted the log G' versus log G" plot in the nematic region farther away from that in the isotropic region, implying that the structure of the sheared specimen was driven to a stronger anisotropic state. It is of interest to observe in Figure 17 that as the applied shear rate is increased from 0.107 to 1.07 s- 1, the log G' versus log G" plot of the sheared specimen moves towards that of the unsheared specimen. We can thus conclude that log G' versus log G" plots are very sensitive to variations, as affected by the extent of steady shear flow, in the morphology of PSHQ10. Notice in Figure 17 that the log G' versus log G" plot in the terminal region of PSHQ10 specimens before being sheared has curvature, closely resembling that observed in the liquid-liquid, two-phase mixtures consisting of a block copolymer and a homopolymer 34'3s. However, the log G' versus log G" plot of a sheared specimen no longer has curvature, resembling that often observed in homogeneous polymers 24---29. This is further evidence that the application of shear flow to PSHQ10 has altered its morphology, producing a more uniform morphology which is close to a monodomain texture, as compared to polydomain textures that existed prior to shearing.

used for each shear rate. Upon cessation of steady shear

C o x - M e r z rule for PSHQIO. It has been reported extensively, since the first study of Cox and Merz4°, that log Ir/*l versus logo9 plots can be correlated with the logarithmic plots of steady shear viscosity (~/) versus shear rate (V). Such a correlation has been obtained primarily f o r homogeneous polymer systems. In the present study, w e investigated whether the Cox-Merz rule holds for PSHQI0 in b o t h t h e isotropic a n d n e m a t i c regions. Figure 18 gives plots of log It/*[ versus log ~o and log r/

measurement began. (~) Data taken using unsheared specimens (i.e. before being subjected to steady shear flow)

versus log~; for PSHQ10 at 190°C, showing that the Cox-Merz rule holds for PSHQ10 in the isotropic region.

to

I

~o-z

I tlllll[ I ~J~J~l J ~JJ,Jl J ~l~n ~o-' io° io' ioz w (rod/s)

Figure 16 Effect of steady shear flow on subsequent complex viscosity of PSHQ10 specimens at 150°C. Oscillatory shear measurements were taken after a specimen was subjected to steady shear flow at shear rates (s-l): ( 0 ) 0.107; (A) 0.536; (ll) 1.07. A fresh

specimen was

flow, the specimen was allowed to rest for 3 h before oscillatory shear

POLYMER, 1994, Volume 35, Number 1

101

Oscillatory shear flow behaviour." S. S. Kim and C. D. Han plots showed a Newtonian plateau region. Thus, by comparing the Ir/*t data obtained from the rest state with the ~/ data obtained in the first shear rate sweep we conclude that the Cox-Merz rule does not hold. However, when comparing the Ir/*l data obtained after the specimen was subjected to steady shear flow with the r/ data obtained in the second shear rate sweep, we can conclude that the Cox-Merz rule seems to hold. In other words,

4xlO4 -

-a =04 - o

o

o

• eo • •

• • •

~o

:

i

i ~llltll

I°lOZ

I

J t ll~Jll

I0-'

I

the C o x - M e r z rule does not hold for a P S H Q l O specimen

i ~llllJ

id

I00 ~" or CO (s-I)

Figure 18 Plots of loglr/*[ versus logco (O), and plots of r/ versus logv (Q) for PSHQ10 specimens in the isotropic region at 190°C

3x=os 105_.Q.. O QO Q Q -

~.

-~~-

_ -~ 4~ -_-__

A •

O Q

~*,QI~

~

CONCLUSIONS

.

A

*

. *

.

• •

-

~.1

[ ~t ~ . 1

Io-~

i

el

lllill

=o-~

:~ or

i J ~t~[~J

Io°

Io ~

co (¢f)

19 Plots oflog l~l*lversuslog coat 140°Cfor a thermallytreated PSHQ10 specimen: (E)) before being subjected to steady shear flow; (O).after being subjected to steady shear flow. Plots of log r/ versus log 7 at 140°Cfor a thermally treated PSQH10 specimen:(A) during the first shear rate sweep;(A) during the second shear rate s w e e p Figure

Plots of log 1~/'1 versus log co and log ~/versus log ~ are given in Figure 19 for a PSHQ10 specimen in the nematic region at 140°C. The thermal history and experimental procedures utilized to obtain the results in Figure 19 are as follows: (1) a specimen, which had received thermal treatment at 190°C, was cooled down slowly to 140°C, and then after a rest for 2 h the specimen was subjected to a frequency sweep; (2) after completion of the frequency sweep, the specimen was subjected to steady shear flow by increasing y from 0.0085 to 0.269 s-~; (3) after a rest for 1 h upon cessation of the previous shear flow, the specimen was subjected to steady shear flow once again by increasing y from 0.0085 to 0.034 s-Z; (4) after completion of the steady shear flow the specimen was subjected to a frequency sweep once again; and (5) finally, the. specimen was subjected to steady shear flow again at y=0.068 s -~. The following observations are worth noting in Figure 19. The shape of the log Iv/*[ versus log co curve after steady shear flow is almost the same as that before steady shear flow, but the values of 1,1"1 after steady shear flow are about an order of magnitude smaller than those before steady shear flow. When the specimen was subjected to steady shear flow for the first time, t/ decreased rapidly with increasing 7, exhi.biting shearthinning behaviour over the entire range of 7 tested which is believed to be due to the presence of polydomains in the specimen. But when the specimen was subjected to steady shear flow for a second time, log ~/ versus log y

102

POLYMER,

1994,

having polydomains, while it seems to hold after polydomains have been broken up under steady shear flOW. This observation is expected to hold for other TLCPs as well. Therefore, we can conclude that the morphological state of a T L C P specimen in the nematic region plays an important role in determining whether or not the Cox-Merz rule holds.

Volume

35, Number

1

In the present study we investigated the oscillatory shear flow properties of an aromatic thermotropic polyester (PSHQ10) in both the isotropic and nematic regions, placing emphasis on the effects of thermal and shear histories. We found that when an as-cast PSHQ10 specimen was annealed at a temperature below its TN[, the rheological properties varied continuously during the experiment and thus it was not possible to obtain meaningful rheological measurements in the nematic region of the TLCPs. We have shown that it would not be possible to investigate the effect of shear history on the rheological properties of a T L C P in the nematic region unless the effect of thermal history is suppressed or isolated. This was made possible in the present study, since the initial morphology of a PSHQ10 specimen was controlled by first heating to the isotropic region and then cooling slowly down to a predetermined temperature in the nematic region. In other words, thermal treatment in the isotropic region followed by a very slow cooling to the nematic region enabled us to control the initial morphology of PSHQ10 specimen in the nematic region. It should be mentioned that earlier Driscoll and co-workers is reported the oscillatory shear flow properties o f a thermotropic polyester, having a chemical structure very similar to PSHQ10, in both the isotropic and nematic regions. However, they did not report the effect of thermal history (i.e. annealing temperature and annealing period) on the rheological behaviour of the polymer in the nematic region. In this study we observed that in the nematic region the PSHQ10 specimen, which had not received thermal treatment in the isotropic state, exhibited a hysteresis effect during oscillatory shear flow, namely the values of Ir/*l for ascending frequency sweep were not the same as those for descending frequency sweep. We found that the application of steady shear flow to a PSHQ10 specimen in the nematic region has a profound influence on subsequent oscillatory shear flow properties. Thus extreme care must be exercised when comparing oscillatory shear flow properties of TLCPs obtained by different research laboratories, unless the shear and thermal histories of the specimens were identical. Finally, we found that the Cox-Merz rule holds for PSHQ10 in the isotropic region and the morphological state of a specimen in the nematic region determines whether or not the Cox-Merz rule holds. We observed that the morphology of PSHQ10 in the nematic region

Oscillatory shear flow behaviour: S. S. Kim and C. D. Han is g r e a t l y i n f l u e n c e d b y the a p p l i c a t i o n of s t e a d y s h e a r

17

flow, which in turn influences subsequent oscillatory

18

shear

flow

properties.

We

thus

conclude

that

the

Cox-Merz rule is not expected to hold when the morphological state of a T L C P is not the same during both steady shear flow and oscillatory shear flow.

19

20 21

REFERENCES 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Jackson, W. J. and Kuhfuss, H. F. J. Polym. Sci., Polym. Chem. Edn 1976, 14, 2043 Wissbrun, K. F. Br. Polym. J. 1980, 13, 163 Prasadarao, M., Pearce, E. M. and Han, C. D. J. Appl. Polym. Sci. 1982, 27, 1343 Simoff, D.A. andPorter, R.S. Mol. Cryst. Liq. Cryst. 1982,110,1 Gotsis, A. D. and Baird, D. G. J. Rheol. 1985, 29, 539 Viola, G. G. and Baird, D. G. J. Rheol. 1986, 30, 601 Wunder, S. L., Ramachandran, S., Cochanour, C. R. and Weinberg, M. Macromolecules 1986, 19, 1696 Wissbrun, K. F., Kiss, G. and Cogswell, F. N. Chem. Eng. Commun. 1987, 53, 149 Zhou, Z., Wu, X. and Wang, M. Polym. Eno. ScL 1988, 28, 136 Irwin, R. S., Sweeny, W., Gardner, K. H., Cochanour, C. R. and Weinberg, M. Macromolecules 1989, 22, 1065 Gonzalez, J. M., Minoz, M. E., Cortazar, M., Santamaria, A. and Pena, J. J. J. Polym. Sci., Polym. Phys. Edn 1990, 28, 1533 Kalika, D. S., Giles, D. W. and Denn, M. M. J. Rheol. 1990, 34, 139 Cocchini, F., Nobile, M. R. and Acierno, D. J. Rheol. 1991, 35, 1171 Guskey, S. M. and Winter, H. H. J. Rheol. 1991, 35, 1191 Driscoll, P., Masuda, T. and Fujiwara, K. Macromolecules 1991, 24, 1567 Wissbrun, K. F. and Griffin, A. C. J. Polym. Sci., Polym. Phys. Edn 1982, 20, 1835

22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40

Blumstein, A., Thomas, O. and Kumar, S. J. Polym. Sci., Polym. Phys. Edn 1986, 24, 27 Sun, T., Lin, Y. G., Winter, H. H. and Porter, R. S. Polymer 1989,3, 1257 Furukawa, A. and Lenz, R. W. Macromol. Chem. Macromol. Syrup. 1986, 2, 3

Driscoll,P., Fujiwara, K., Masuda, T., Furukawa, A. and

Lenz, R. W. Polym. J. 1988, 20, 351 Driscoll, P., Masuda, T., Furukawa, A., Lenz, R. W. and Bhattacharya, S. Polym. J. 1990, 22, 609 Kim, S. S. and Han, C. D. Macromolecules 1993, 26, 3176 Ferry, J. D. 'Viscoelastic Properties of Polymers', 3rd Edn, Wiley, New York, 1980 Han, C.D.andJhon, M.S.J. Appl. Polym.Sci. 1986,32,3809 Han, C. D. and Lem, K. W. Polym. Eng. Rev. 1983, 2, 135 Chuang, H.K.andHan, C.D.J.Appl. Polym.Sci. 1984,29,2205 Han, C.D.andChuang, H.K.J.Appl.Polym.Sci. 1985,30,443 Han, C.D.andYang, H.H.J.Appl. Polym. Sci. 1987,33, 1199 Han, C. D. J. Appl. Polym. Sci. 1988, 35, 167 Han, C. D. and Kim, J. J. Polym. Sci., Polym. Phys. Edn 1987, 25, 1741 Aoki, Y. Rheol. J. (Japan) 1979, 7, 20 Han, C.D.,Kim, J.andKim, J.K.Macromolecules1989,22,383 Han, C. D., Baek, D. M. and Kim, J. K. Macrornolecules 1990, 23, 561 Kim, J., Han, C. D. and Chu, S. G. J. Polym. Sci., Polym. Phys. Edn 1988, 26, 677 Han, C. D., Kim, J., Baek, D. M. and Chu, S. G. J. Polym. Sci., Polym. Phys. Edn 1990, 28, 315 Onogi, S. and Asada, T. in 'Rheology' (Eds G. Astarita, G. Marrucci and L. Nicolais), Vol. 1, Plenum Press, New York, 1980, p. 127 Lin, Y. G. and Winter, H. H. Macromolecules 1988, 21, 2439; 1991, 24, 2877 Kim, S. S. and Han, C. D. J. Rheol. 1993, 37, 847 Grizzuti, N., Cavella, S. and Cicarelli, P. J. Rheol. 1990, 34,1293 Cox, W. P. and Merz, E. H. J. Polym. Sci. 1958, 28, 619

POLYMER, 1994, Volume 35, Number 1

103